Smart Meter Data based Load Analysis Using Clustering Technique

Abstract

In this study, clustering is investigated as a method for improving energy efficiency based on smart meters for a number of household applications that are both currently available and expected to become available in the near future. These applications include smart thermostats, smart lights, smart water heaters, smart washing machines, and smart refrigerators. We describe a novel approach to load balanced clustering that is founded on the K-means Clustering algorithm. Our algorithm's major goal is to optimize network lifetime while maintaining acceptable sensing coverage in scenarios in which sensor nodes generate either uniform or non-uniform data traffic. This can be accomplished by maintaining acceptable sensing coverage. We also provide a new clustering cost function that takes into consideration not only the volume of traffic but also the amount of work that is required to communicate across substantial geographic distances. This is done so that we can achieve this objective. We demonstrate that our algorithm is able to improve both load analysis as well as load balancing in the domestic area by running extensive simulations that compare the proposed algorithm to leading state-of-the-art clustering approaches and then comparing the results to one another. This allows us to demonstrate that our algorithm is able to improve both load analysis as well as load balancing in the domestic area. In addition to this, it demonstrates the culmination of a stage in the processing of a dataset in order to compute the typical quantity of energy load that is utilized by consumers.